The evolution of computer conferencing may be far from complete, moreover. There are undoubtedly many refinements to be made to existing rules and genres. There are probably new genres that remain to be developed. There are almost certainly effects that have yet to be observed and characteristics that remain to be identified. Most important, perhaps, are the mediators that remain to be applied to computer conferencing. There are many such mediators on the horizon. Some can be reasonable projected on the basis of the problems of computer conferencing as it exists now. Others represent technological innovations that offer the possibility of significantly enhancing and/or extending the possibilities inherent to a computer conference.
The leading problems with using IBMPC include unrestrained discussion and tight reviewing. One sees no satisfactory solution to either of these problems on the horizon. Stricter rules and tighter enforcement of those rules could constrain discussion, but only at the expense of a perception of tighter reviewing. Looser enforcement could reduce the perception of tight reviewing, but only at a cost of even less constrained discussion. Either action could substantially increase the workload of the IBMPC administrators, for whom the reviewing of IBMPC is only a part time job (they develop software as a full time job).
It may be important, relative to any solutions, to better understand those segments of the IBMPC community that are concerned with each of these conflicting problems. The individuals who complain most about unrestrained discussion may consistently be among the less frequent contributors to IBMPC. The individuals who complain most about censorship may be among the more persistent appenders. Alternatively, individuals who complain about unrestrained discussion may be more concerned with the information value of IBMPC while those concerned with censorship are more concerned about the discussion value of the facility. Confirmation of any of these possibilities might at least inform the alternatives for future action.
One such alternative might include setting up an information oriented read-only subset of IBMPC that excludes strictly discussion oriented appends. Whole forums, including IBMPC's meta-discussions and lightening rods, would be excluded from such a facility. Appends to the remaining forums would only reach this facility after having been appended to IBMPC and designated by an editor as belonging on the subset facility. The result would be a sort of "IBMPC abstracts". Experiments in this direction can be found in NEWSXPRT FORUM on IBMPC, which extracts the really useful information appended to NEWSCLIP FORUM. This alternative might be particularly attractive (given adequate funding) if unrestrained discussion is largely a problem among the less frequent contributors to IBMPC or those concerned largely with the information value of IBMPC.
A more labor intensive solution might entail formally abstracting the contents of IBMPC into technical reports. This has been attempted before on a voluntary basis. It might be more successful if it was the job of a group of full time technical writers to abstract the useful technical content of IBMPC into technical reports on an ongoing basis. This solution is probably unworkable as it would require a formal investment in internal conferencing by IBM, something that doesn't appear to be on the immediate horizon. Note that the difficulties in making this solution work also decrease the likelihood that solution to the need for better forum summaries will be found.
An intermediate, but still rather labor intensive, solution, might be found in expanding the role of answers files. ANSWERS files currently used to house commonly asked questions and their answers. A more aggressive answers files program would encourage forum owners to move any really valuable information to an associated ANSWERS file. A solution like this one is being tried on another conferencing facility in IBM as this is being written. It should be interesting to see the results of the experiment, which has so far been highly controversial.
Less labor intensive solutions might be found in new front end options. One such option might allow readers to skip whole threads of discussion (based on the subject header) with a keystroke. Another such option might require contributors to designate appends as containing information or discussion. Front end options (including mail filters) could then sort out the information and discussion appends according to the preferences of the reader. Those who prefer to only see the information on IBMPC could then readily skip or filter out discussion. This designation would probably lead to informal rules concerning what constitute each that would evolve as information only readers of IBMPC complained about mislabeled appends.
Any of these solutions might also further reduce the problems of information overload and quality of information on IBMPC. None of these solutions will improve the problem of perceived censorship on IBMPC. Given the constraints under which IBMPC must be administered, there is probably no solution to this perception. This is unfortunate, as ill-timed accusations of censorship probably hinder open discussion on IBMPC to a greater extent than they help.
"Ill-timed" is key here, for there are any number of occasions where such accusations have directly followed the expenditure of considerable effort on the part of IBMPC's administrators and forum owners to keep a controversial topic open. The effect, for some reviewers, has been to expend less effort on subsequent occasions. It is probably to be easier to take the heat for censorship when one has made little personal investment in the maintenance of a borderline topic.
Several of the solutions outlined above might also simplify the task of finding things on IBMPC. The ultimate solution, however, is probably some sort of intelligent keyword search engine that can take a request for information and resolve it into a list (preferably a moderately short one) of possibilities. Experimentation with the use of a simple key word search engine on another conference in IBM has not lived up to the implementer's expectations. Further experimentation with such search engines can be expected, but no real solution can be expected until the search request is built directly into the conferencing software and the user front ends used to access the conference. Moving the search request into the conferencing software would allow such requests to be handled in exactly the same way requests are handled now. Moving the search request interface into user front ends would give the capability exposure that it would not have otherwise.
Incorporation of such facilities into user front ends is only one of many things that can be done to improve the user interface to computer conferencing within IBM. The next big step in such interfaces will probably be movement of the interface from the host terminal screen to the more flexible personal computer full screen interface. Experiments with this kind of interface are already underway under OS/2 Presentation Manager. The big advantages of such interfaces can be found in faster editors that allow users to read and create appends more quickly, menu and button driven visual interfaces that offer users more options and make them easier to find, and context sensitive direct manipulation interfaces that allow the extension of existing computer conference information in the direction of hypertext.
Another major direction for computer conferencing user interfaces will be some level of intelligent processing of forum content. Today's IBMPC participant cannot follow every forum, but may have an interest in topics that that may appear in forums the user doesn't normally follow. User front ends that can scan the forums a user doesn't follow for topics of interest extends the users ability to participate on IBMPC. This kind of intelligent conference profiling might also potentially be used to automatically skip uninteresting topics within the forums one normally viewed.
Other possible innovations that may change the shape of computer conferencing in the foreseeable future are related less to the current problems of IBMPC than to the technical capabilities that will be available for use in future computer conferencing systems. There is nothing driving computer conferencing to make use of these innovations except their possibility. Hence none is as likely to succeed as some of the solution oriented innovations outlined so far. Each offers computer conferencing interesting new possibilities, however. These possible innovations include:
Voice to text interfaces, in which the computer converts the spoken word to written text, offer the potential to significantly increase text entry speeds, quite possibly changing the dynamics of computer conferencing in the process. Even at speeds of 40 words per minute (much slower than we typically talk), voice to text interfaces should make it much easier to composed appends to computer conferences. One result, one expects, will be increased participation in computer conferences. A second result might well be a more casual style of interaction.
Prototype hypermedia computer conferencing front ends allow the interactive construction of hypertext links and the integration of text, graphics, signatures, animations, music, and working programs. It remains to be seen, however, how extensively these features will be used. Computer conference append creation is already comparatively slow. The addition of graphics, music and other elements can only slow it further, perhaps without adding much to the interaction. One expects such features will be used. One expects such features will prove worthwhile under some circumstances. But one expects that even when such features are readily available for use, that most appends to computer conferences will contain only text.
None of these potential innovations solve the problems that computer conferencing will have to overcome if it is to become highly successful outside of corporate environments or network environments like BITNET. Vendors like Prodigy are beginning to address these problems with solutions like affordable fixed price network connections via local telephone connections. Vendors throughout the PC industry are addressing a second problem with affordable low cost personal computers and software.
The medium appears to be well positioned for success from a typological perspective. It is clearly differentiated from other media in the formal and user typologies. Hence if the hypotheses concerning typological distinction and media success are correct, there appears to be an excellent possibility that computer conferencing will face little competition in the media marketplace.
This position is supported by the distinctive genres of computer conferencing that have been observed in this study. Computer-mediated query, interactive software development, meta-forums, electronic focus groups, and computer mediated memorial services all offer distinctive possibilities for interaction that are not as readily achieved in the genres of other media. These distinctive genres offer the possibility that computer conferencing will be used outside of IBM for the same kinds of applications that it is used for inside IBM; that the medium will find large niches in the general marketplace that simply are not filled by other media.
This possibility appears to be supported by the success computer conferencing appears to be enjoying in the mass market already. Many personal computer users are already making some use of computer conferencing through services like Prodigy, CompuServe, BIX and others. The number of such users of various computer conferencing services can be expected to grow steadily, moreover, as the number of people using computers increase and fixed rate computer conferencing becomes available in more places. The percentage of computer conferencing users in the general population is small by comparison with the near universal electronic mail coverage and extremely high IBMPC participation inside IBM. Still, the number is high, and services like CompuServe now claim more regular users than IBM has employees.
If the experience of computer conferencing in IBM is any indication, that number can be expected to grow substantially in the future. Computer conferencing is now one of the leading forms of communication within IBM. The medium's use by individual IBMPC participants was comparable to that of group meetings (a major activity in IBM) and radio listening in 1988 and had grown rapidly since 1986. IBMPC's growth statistics also support this notion, indicating that IBMPC may have been read by between one fifth and two fifths of all IBM employees in 1988; indicating that the rapid growth that IBMPC has enjoyed since 1985 may continue today.
One concludes, then, that the prospects of computer conferencing's success are excellent. Computer conferencing has not achieved the lofty expectations of Hiltz and Turoff (1978) projected for 1988 outside IBM. We are not yet a "network nation". Those expectations have been more than satisfied within IBM, however, where computer conferencing has turned IBM into a global "network corporation" where employees interact with transoceanic colleagues about as casually as they interact with their colleagues on the same aisle. The same kinds of growth that were observed early in the history of IBMPC are now being observed in the interaction on computer conferencing facilities on BITNET, with an international community of scholars beginning to interact in ways that has never been possible before. It is reasonable to expect, moreover, that this kind of interaction will, probably within the next ten to twenty years, become readily available to, at the very least, most Americans, most Japanese, and many Europeans.
The review of literature documents several deficiencies in the prior computer conferencing literature. This study has attempted to remedy these deficiencies in several ways.
First, it observed a large scale and rapidly growing computer conferencing facility for a period of over six years. The opportunity to study the IBMPC computer conferencing facility, and in particular the opportunity to study the decisions of the people who shaped IBMPC as they were being made, must ultimately be regarded as a fortuitous accident. Still, the author had been actively searching for a study target for over two years when the opportunity presented itself.
Second, the study purposely avoided making projections of its own until the very end. It is only in the last several pages that we have attempted to guess at either future directions in computer conferencing or the medium's prospects for success outside IBM. We have sought to observe rather than project under the general supposition that the existing base of observation remained too thin to allow theoretical assumptions to be made about computer conferencing. A major goal of the study was to create a base of data from which theory might be generated.
Third, the study has generated, on the basis of its observation, a theory of media evolution that integrates interpersonal and mass media in a single theoretical framework. This theoretical framework, which is argued on the basis of observation of other media, is built on insights generated in intersection of the different methods and perspectives, as seen below.
Fourth, the study has persistently sought to apply a variety of observational methods and perspectives to the study. The principle method of the study is participant observation, but the results of the participant observation has been repeatedly tested using other methods, including surveys, archival data, and theory. The mixing of these diverse methods appears to have allowed diverse methods to inform each other, as is seen in the following examples:
These are but a few of the more important examples of the use of participant observation, archival, survey, and other data in conjunction with theoretical justifications as this study sought to cross check and enhance its insights.
This study has some clear limitations:
This study started with two themes. One, the mechanisms by which media evolve, has been explored thoroughly in both the theoretical chapters which initially presented these mechanisms and the observational chapters in which these mechanisms can be readily observed. These mechanisms appear to be well demonstrated in both the detailed examination of computer conferencing and the casual consideration of many other media. Hence one might readily conclude from this study that there may well be something in the grammar, typologies, and theory of media that are presented here.
The second, the nature of computer conferencing, has also been explored thoroughly. Although observation has been restrained to a single computer conferencing facility, IBM's IBMPC computer conferencing facility, interaction on the facility has been explored from many perspectives. We have considered its mediators, characteristics, genres, rules, problems, impacts, and evolution using a variety of tools that cross checked and extended our observation of the nature of computer conferencing in many ways.
While we cannot reasonably state that the second theme has been demonstrated, it should be understood by now that, at least in the case of IBMPC:
All of these observations tell us a great deal about the nature of computer conferencing and the reasons for its success.
Still, it would appear that there is no "nature" of computer conferencing. The great lesson of the first question and its associated theory of media is that nothing about the medium would appear to be in any way natural or inherent. Computer conferencing is an invented medium, both in the software used to access the facility and the control of such access; both in the rules that restrain interaction and the genres that encourage it. The result is an artificial and highly malleable medium that can be readily adapted to many purposes.